II. Protein folding and unfolding

A. Protein quality control

The maintenance of cellular functions relies on a correctly folded proteome. Indeed, the mechanism by which protein homeostasis (or proteostasis) is ensured requires a well-coordinated control of protein synthesis, folding, conformational integrity, and protein degradation. In this network,

which coordinates these processes, chaperone-assisted protein folding and functional protein degradation pathways are of particular importance. During her lecture, Dr Bertolotti highlighted that protein misfolding diseases have a late onset and that the pathological deposition of aggregated misfolded proteins is a feature of many diseases, such as Alzheimer disease, Huntington disease, and diabetes.
In particular, in obese patients, the efficiency of the proteostasis network is altered and insufficiently capable of adapting to metabolic stress. When insulin production needs exceed β-cell capacity, chaperone production is impaired, and proteostatic collapse occurs, resulting in an accelerated development of overt diabetes,26 which is why “assisting” a malfunctioning protein quality control system could be a promising strategy in the prevention of T2D (Figure 4).27The main players in protein quality control systems are chaperones and protein degradation systems: the ubiquitinproteasome system and autophagy28 (Figure 5). During times of high protein synthesis needs, the ER has to cope with the high demand in protein folding that leads to the unfolded protein response, a coordinated translational and transcriptional program aimed at adjusting protein quality control.

Briefly, protein misfolding stress in the ER activates three different pathways: (i) protein kinase R–like endoplasmic reticulum kinase (PERK) phosphorylates eukaryotic initiation factor 2α (eIF2α), leading to general translational inhibition and transcriptional activation of
specific target genes (eg, those encoding protein phosphatase 1 regulatory subunit 15A [PPP1R15A] and CCAAT/enhancerbinding protein [C/EBP] homologous protein [CHOP]); (ii) inositol-requiring enzyme 1 (IRE1) mediates X-box binding protein 1 (XBP1) splicing and the transcriptional activation of chaperones; and (iii) activating transcription factor 6 (ATF6) translocates to the Golgi apparatus and converts into a transcription factor to
activate unfolded protein response genes. Interestingly, protein misfolding also occurs in the mitochondria and induces an unfolded protein response, enhancing transcription of mitochondrial chaperones in the nucleus.28 By using proteasome inhibitors, Dr Bertolotti pointed out that proteasome degradation is crucial for cell viability by maintaining an amino acid pool size that is required for proper proteostasis. Indeed, the in vitro administration of both proteasome inhibitors and amino acids is viable.28,29

B. Insulin toxic misfolding

As shown in Figure 3, insulin synthesis is initiated as preproinsulin, and several steps are required to ensure proper proinsulin generation and folding: (i) correct orientation of the signal peptide; (ii) translocation into the ER; and (iii) signal peptide cleavage. The efficiency of proinsulin folding has been remarkably preserved during evolution; however, proinsulin misfolding could occur because of very high proinsulin biosynthetic load in the ER due to increased needs as a result of genetic mutations altering the proinsulin coding sequence or by affecting genes in the ER folding environment.30 Importantly, proinsulin misfolding has been described and linked to a defect in insulin production in rodent models (eg, Mody mice having a mutation in the insulin 2 gene) and human patients (eg, with a mutant INS-gene induced diabetes of youth [MIDY]).31,32 Furthermore, proper regulation of the unfolded protein response is essential for the well-being and function of β-cells, which is exemplified in monogenic forms of diabetes, resulting from mutations in unfolded protein response genes.33

Figure 634 presents the structure of insulin as it was defined in 1969.35 As explained by Dr Lawrence, the aromatic triplet is thought to play a critical role in the conformational changes associated with the reorganization of the insulin receptor ectodomain upon hormone binding.36 In this regard, single-particle, cryo-electron microscopy has allowed significant advancements,37 which paved the way for the design of novel therapeutic “insulin-like” molecules that can activate the insulin receptor.38

It is now well established that an intact unfolded protein response is necessary for β-cell function and therefore glucose homeostasis; several mouse models of T2D display increased expression of unfolded protein response genes and oxidative stress genes in their islets.39 Autophagy is also essential for proper islet architecture and function; for example, mice deficient for Atg7, a mediator of autophagy, in β-cells have an impaired glucose tolerance with reduced insulin secretion.40,41As described in Figure 5, the unfolded protein response leads to a decrease in general protein translation rates, increased production of ER chaperones, and the ubiquitinylation of newly synthetized proteins and their degradation in the proteasome, this last part being called “ERAD” for ER-associated degradation pathway. As discussed by Peter Arvan, several studies implicate a role for the transcription factor ATF6α in controlling β-cell function and survival.42

Briefly, the XBP1 and ATF6 pathways maintain a productive ER protein folding environment and increase the ERAD to degrade misfolded proinsulin,43,44 thus helping maintain a population of mature and well-differentiated β-cells. In addition, Dr Leibowitz explained that an accumulation of misfolded proinsulin in the ER stimulates the unfolded protein response, which may eventually lead to apoptosis through a process called the terminal unfolded protein response.33 Taken together, this data shows that the restoration of a proper unfolded protein response in pancreatic β-cells should protect against the development of overt diabetes; interestingly, preliminary evidence has been obtained in mice treated with a chemical ER stress mitigator, the bile acid tauroursodeoxycholic acid (TUDCA).44

Importantly, the actions of TUDCA were dependent on ATF6 and were lost in mice with a β-cell–specific deletion of ATF6.

The unfolded protein response is also activated in response to inflammatory signals, and Bart Roep, in his elegant lecture, revealed that this adaptive response may play a deleterious role in autoimmunity and β-cell death by generating neoantigens. Indeed, proinsulin is targeted to the secretory pathway in normal conditions; however, under stress and activation of the ubiquitin proteasome system, small new peptides are generated and targeted to the cell surface and presented as antigens to the host immune system, thereby eliciting potentially potent T-cell mediated responses (Figure 7).45 In addition, work from Eizirik et al provided evidence that, under inflammatory stress conditions, proinflammatory cytokines increased and affected gene expression, leading to increased alternative mRNA splicing events and generation of β-cell neoautoantigens.21